Whether the design system's tokens are actually used — or whether hardcoded colors, spacing, and type have crept back in and drifted the UI out of sync.
Free & open · no signup · read-only — it ends by asking · nothing leaves your machine
You are working inside this repo. Mission: check whether the design tokens are the single source of truth they claim to be — or whether hardcoded values have crept in, quietly drifting the interface out of alignment one one-off at a time.
This assumes a token system exists and audits whether it is used. No system yet? Run 15 to derive one from the drift.
Read-only pass. Read the token definitions and the styles that should consume them; change nothing but the report file.
Phase 2 sweeps the codebase through every one of these, citing file and line for each finding.
hex/rgb literals where a color token belongs; near-duplicate one-off colors
magic pixel values instead of the spacing scale
ad-hoc font sizes, weights, and line-heights outside the type scale
surfaces (states, dark mode, borders, shadows) with no token, so people improvise
the same value defined in several places that can fall out of sync
whether tokens flow from one definition or are copied per surface
is there a lint rule or check, or does drift only get caught by eye
One structured report at the repo root — or in reports/, if you keep one — the same shape every time, ready for a teammate — or the optional Studio — to act on.
Every Goal Prompt follows the same four steps, so results are consistent and repeatable — no matter which one you run.
Locate the token definitions.
Cite file and line for every finding.
Rank by visible drift and spread.
Create TOKEN-ADOPTION.md at repo root.
Copy it and paste it into your agent inside the repo you want checked.
Install the goal plugin once — two commands — then just type /goal:design-token-adoption.
/plugin marketplace add GhostlyGawd/goal-prompts/plugin install goal@goal-promptsOr install only this Goal Prompt as /goal-design-token-adoption:
curl -fsSL https://goal-prompts.vercel.app/install | BRIEF=105 shLet an agent fetch it mid-conversation, or pull the raw Goal Prompt by URL.
https://goal-prompts.vercel.app/raw/105.mdNothing hidden — this is the whole Goal Prompt, verbatim. Read it in a minute, edit it, or copy it as-is.
# Goal: Design-Token Adoption Audit You are working inside this repo. Mission: check whether the design tokens are the single source of truth they claim to be — or whether hardcoded values have crept in, quietly drifting the interface out of alignment one one-off at a time. This assumes a token system exists and audits whether it is used. No system yet? Run 15 to derive one from the drift. Read-only pass. Read the token definitions and the styles that should consume them; change nothing but the report file. ## Phase 1 — Find the source of truth - Locate the token definitions: color, spacing, type, radius, shadow, motion. - Note how they are meant to be consumed and where they are defined more than once. - Scan the styles for literal values that should reference a token. ## Phase 2 — Audit through 7 lenses Cite file and line for every finding. 1. **Color creep** — hex/rgb literals where a color token belongs; near-duplicate one-off colors 2. **Spacing creep** — magic pixel values instead of the spacing scale 3. **Type creep** — ad-hoc font sizes, weights, and line-heights outside the type scale 4. **Token coverage** — surfaces (states, dark mode, borders, shadows) with no token, so people improvise 5. **Duplication & drift** — the same value defined in several places that can fall out of sync 6. **Single source of truth** — whether tokens flow from one definition or are copied per surface 7. **Enforcement** — is there a lint rule or check, or does drift only get caught by eye ## Phase 3 — Curate - Rank by visible drift and spread: a one-off color repeated across many components outranks a single stray value. - For each, map the literal to the token it should use, or the token the system is missing. - Separate "use the existing token" from "the system lacks a token"; both cause creep. ## Phase 4 — Report Create `TOKEN-ADOPTION.md` at repo root: 1. **Adoption score** — per category (color, spacing, type): tokenized versus hardcoded 2. **Offenders** — the worst hardcoded values with locations and the token they should use 3. **Missing tokens** — the surfaces with no token, so improvisation is inevitable 4. **Enforcement** — the lint rule or check that keeps hardcoded values from returning Start the report with today's date. If `TOKEN-ADOPTION.md` already exists from a previous run, read it first and lead with what changed since. ## Rules - One value, one token, one definition; a copied value will drift - A missing token is a cause of creep, not just the literal that fills it - No styled UI in this repo? Say so in a one-paragraph null report and stop — a null result is a valid finding. - If a `reports/` directory exists at the repo root, write the report there instead of the root. - Before asking, present the top findings as a ranked list in plain words - Report only — end by asking which drift to correct first
reports/ directory exists at the repo root, write the report there instead of the root.Curated neighbors — briefs that answer the adjacent question, worth running in the same session.
Census the drifted colors, spacing, and duplicate components; propose the token set and migration order to unify them.
The palette as it actually ships — every hex in the codebase, contrast ratios computed, semantic drift named, and whether dark mode is a theme or an apology.
Where the eye actually goes versus where it should — focal points, size and weight and contrast doing real work, and the screens where everything shouts at once.
The type scale, or the lack of one — sizes, weights, line lengths and heights, font loading, and every place text is asked to do a job at the wrong size.
The palette as it actually ships — every hex in the codebase, contrast ratios computed, semantic drift named, and whether dark mode is a theme or an apology.
The spacing scale nobody wrote down — margins and paddings inventoried, grid discipline checked, alignment breaks found, and density judged screen by screen.